Debated in Parliament on 8 Nov 2022.
Minister for Communications and Information.
Mr Speaker, I beg to move, "That the Bill be now read a Second time."
Sir, a study conducted by the Ministry of Communications and Information (MCI) in June this year found that almost 80% of Singapore residents are concerned with online harms. In stark contrast, when we ask people how they feel about walking the streets of Singapore alone at night, 97% said that they would be comfortable to do so. There is obviously a sizeable gap between how safe Singaporeans feel online and offline.
Today, most of us remain connected online throughout the day. Online services have become the key conduits through which we communicate and consume content. Because of this, the prevalence of harmful online content on these services can have negative serious consequences on the physical, emotional and mental well-being of society.
The Bill we are debating today is not the first law introduced to secure our online space. The Government has, over the years, introduced targeted laws to deal with specific types of harmful online content and behaviour, including:
(a) Falsehoods, which are dealt with under the Protection from Online Falsehoods and Manipulation Act, or POFMA.
(b) Foreign interference, which is addressed under the Foreign Interference (Countermeasures) Act, or FICA.
(c) Online harassment, such as cyberbullying, which is dealt with under the Protection from Harassment Act, or POHA. POHA was also recently updated in 2019 to cover doxxing.
Our laws have served to protect many Singaporeans. POFMA was integral to Singapore's response to COVID-19, allowing the Government to address the deluge of misinformation which made COVID-19 not just a pandemic, but an "info-demic".
Recently, The Straits Times reported a record high number of protection orders filed and granted in 2021 under POHA, more than double the number in previous years. Lawyers attributed the spike in applications to media attention on the issue of harassment. The application process has also been enhanced with the opening of the Protection from Harassment Court.
However, there are still gaps that need to be addressed. One growing concern is content encouraging suicide and self-harm. Just two months ago, an investigation in the United Kingdom (UK) concluded that 14-year-old Molly Russell took her own life after being exposed to thousands of self-harm and suicide related posts in the months leading up to her death. Many of these posts portrayed suicide as an inevitable consequence of depression.
There have also been reports of users' accidental deaths while attempting to mimic videos of impossible physical stunts. Unknown to some victims, these reckless acts and dangerous challenges had been heavily edited.
Our children, who may lack the capacity or maturity to deal with certain types of content are particularly vulnerable when exposed to inappropriate content and unwanted social interaction online. In June this year, MCI conducted a study which asked respondents what online content they felt children needed to be most protected from. The top three were sexual content, cyberbullying and violent content.
In a dialogue with youths held by MCI and the National Youth Council last year, participants indicated that the top three online harms they and their peers faced included being insulted online, impersonated by someone else and receiving unwanted contact from another person.
If such harmful content existed only on websites, the Infocomm Media Development Authority (IMDA) would be able to deal with them under the existing Broadcasting Act. But today, users are much more likely to consume content from the feeds of social media services, where such harmful content can be pushed via algorithms, and spread quickly through our social connections.
Just two weeks ago, Meta announced that Facebook recorded nearly 2 billion daily users, while Instagram recorded 2 billion active monthly users. TikTok has been downloaded over 3.5 billion times worldwide since its launch, while YouTube recorded 30 billion daily views on "YouTube Shorts".
As the Internet evolves, so must our laws. In the book "Tools and Weapons" co-authored by Microsoft's President Brad Smith and Carol Ann Browne, the backwardness of some cybersecurity measures was likened to "digging trenches to defend against missiles".
In the same way, we must recognise that online content can inflict serious damage on our people and communities, if our laws fall short. We must have the ability to deal with harmful online content accessible to Singapore users, regardless of where the content is hosted or initiated.
The entities controlling the biggest and most popular online communication services (OCSs) or platforms accessible in Singapore all operate outside of Singapore and fall outside the legal remit of the Broadcasting Act today. To ensure that Singapore users of these services and platforms can be kept safe, we must be able to take appropriate action on these entities, as long as they provide content accessible by Singapore users.
We are not alone in thinking this way. There is a growing consensus that rules must be put in place to prevent harms in the online world, just as in the physical world. Calls for online services to take greater responsibility in ensuring safety on their platforms have also led to jurisdictions such as the UK, the European Union (EU), Germany and Australia to introduce or propose new online safety laws.
Mr Speaker, I seek your permission to distribute handouts to the Members, which summarise online safety laws enacted or under consideration in these jurisdictions.
Please proceed. [Handouts were distributed to hon Members.]
Mr Speaker, the Singapore public, like many other societies, is concerned over the potential damage caused by harmful online content and expects social media services to take greater responsibility to protect their users.
In July and August this year, MCI conducted a public consultation and series of engagements on our proposals to combat harmful online content and received more than 600 responses. Respondents expressed the desire for safety features to manage their exposure to certain types of content.
Similarly, MCI's June 2022 study, found that nine in 10 respondents felt that such measures would protect users from harmful online content to at least a moderate extent. Parents, in particular, were concerned over viral social media content which featured dangerous pranks and challenges, harmful advertising, cyberbullying and explicit sexual content. Some suggested keeping younger users in mind when developing safety features, including tailoring content moderation thresholds and ensuring young users can easily report inappropriate content.
In a separate poll conducted earlier this year by the Sunlight Alliance for Action, Singaporeans ranked reporting systems and laws to tackle online harms as the top two measures that would facilitate help-seeking.
I will now explain our approach to enhancing online safety for Singapore users and Members will find that there are similarities to the practices elsewhere, examples of which I have circulated.
The first is to tackle the problems in an accretive manner. Rather than take a "Big Bang" approach which some countries are attempting and have an all-encompassing law, let us design our laws in a considered and calibrated manner.
[Deputy Speaker (Ms Jessica Tan Soon Neo) in the Chair]
Second, as far as possible, be outcome driven instead of being overly prescriptive. In today's context, we are dealing with a vast volume of user-generated content. Rather than chasing individual pieces of content, we must ensure that systems and processes to regulate the content are put in place and maintained by the platforms. Instead of prescribing how these systems and processes are set up, we should specify the outcomes they ought to achieve.
The third, and perhaps, most important of all, is to recognise that laws are not a silver bullet. The Government will need to work with partners, including our citizens, to tackle harmful content and enhance the safety of users online.
Today's online content service providers are different from traditional local broadcasters and require a different regulatory approach. In fact, each type of service is different. "Social media services" are not the same as "over-the-top media services", which also operate differently from "game distribution services".
The Bill allows us to adopt this accretive approach by building on existing laws to introduce new ones; so that over time, our foundations for digital safety become stronger. If passed by Parliament, this Bill will create a new part in the Broadcasting Act to regulate "online communication service", which are electronic services that enable users to access or communicate content via the Internet.
The regulations will only apply to specified types of "online communication service", which are listed in a Schedule under the Broadcasting Act.
For now, we will only specify one type of OCS in the Schedule and that is "social media services". Under the Bill, a social media service is defined as an electronic service, whose sole or primary purpose is to enable online interaction or linking between two or more users, including enabling users to share content for social purposes; and which allows users to communicate content on the service.
Why have we chosen to regulate social media services as a matter of priority? Well, because three in five users or thereabout from MCI's June 2022 survey experienced harmful content online using social media platforms. This is the highest proportion compared to other platforms, such as e-commerce sites, search engines and news sites.
Given the voluminous user-generated content in today's evolving online space, it is not efficient to regulate individual pieces of content.
IMDA will instead focus on system-wide measures which are more effective at scale.
Under sections 45K and 45L of the proposed Bill, IMDA will be able to designate OCSs with significant reach or impact in Singapore, and require them, via the Codes of Practice, to put in place measures to keep Singapore users safe.
This approach is similar to how we go about regulating fire safety. Building owners, occupiers and qualified persons must adhere to the Singapore Civil Defence Force (SCDF)'s Fire Code, which requires them to put in place systems and processes to maintain high fire safety standards, to keep their occupants safe.
Likewise, OCSs must have in place systems and processes to minimise Singapore users' exposure to, and mitigate the impact of, harmful content on their platforms. IMDA will impose these requirements on designated OCSs via Codes of Practice. By stating in the Codes the outcomes which regulated services must meet, IMDA aims to provide sufficient clarity on what the services must do to protect users, whilst allowing some flexibility for them to adjust their approaches.
We can also expect IMDA to update the Codes from time to time. This will allow us to be agile and responsive to technologies as they evolve.
But before introducing new requirements, IMDA will consult and work collaboratively with service providers to assess the most suitable approaches to strengthening safety on their platforms.
Under the Bill, IMDA does not have unfettered ability to issue new Codes. The new section 45L sets out that IMDA can issue Codes for the following purposes:
First, to ensure services have systems or processes in place to address harmful content.
Second, to provide practical guidance or certainty in respect of what content should be covered.
Third, to set out the procedures that service providers must follow when audits are carried out.
Fourth, to require services to collaborate with approved researchers to understand systemic risks relating to the service.
Earlier, I explained that we will apply our new laws to social media services as the first type of OCS. Let us now turn to the Code that designated social media services with significant reach or impact in Singapore must comply with.
In October, IMDA issued a draft copy of the "Code of Practice for Online Safety". This draft Code comes after an extensive study of international online safety legislation as well as proposals and engagements with major social media services in Singapore, including Facebook, YouTube, Instagram, TikTok, Twitter and HardwareZone.
The social media services consulted were receptive to the proposals laid out in the draft Code and the Bill. They support the Government's commitment to find innovative and effective solutions to combat harmful online content and recognise the need to improve online safety.
The designated social media services will be expected to meet the key outcomes as follows:
First, minimise Singapore users' exposure to harmful content and empower users with tools to manage their own safety. The social media services must also take additional steps to minimise children's exposure to inappropriate content and provide tools allowing children or their parents to manage their safety.
Second, make available an easy-to-use mechanism for Singapore users to report harmful content and unwanted interactions.
Third, provide transparency on the effectiveness of their measures in protecting Singapore users from harmful content. Designated social media services must provide information that reflect Singapore users' experience on their services. This will allow users to make informed decisions about how they use the service.
If the Bill is passed, IMDA will further consult relevant social media services, before finalising the Code for issuance.
We believe that the Code of Practice for Online Safety will reduce users' exposure to harmful online content, but it will not eliminate them completely. Part of the reason is that these social media services tend to operate globally, drawing in users and content from around the world. Their safety measures are not tuned to reflect an in-depth understanding of Singapore's local context or our racial and religious sensitivities.
Members may recall that in the early days of the COVID-19 pandemic, supermarkets were purportedly running out of toilet paper. A social media post surfaced, suggesting that people use the Bible and the Quran as toilet paper. This post was religiously very offensive and denigrated two religions in Singapore. However, it was not moderated nor removed by the platform concerned. IMDA had to step in to engage the platform and the platform eventually disabled access to the post.
There may also be egregious content on non-designated social media services, which are not subject to the Code of Practice for Online Safety. In May last year, a poll published on a social media service sexualised local female Islamic teachers, asked users to rank them and further promoted sexual violence against them. The post went viral and the modest reach of this particular service received a sudden big boost. It not only caused great distress to the individuals involved, but also unsettled many others in the community.
These issues are like fires that occur, even as the Fire Code has prevented most fires. In such instances, we must have firefighters who are properly equipped to act quickly, so as to minimise, if not prevent, serious injury and damage.
If Parliament agrees, the new section 45H proposed by this Bill will allow IMDA to act as an "online firefighter", to direct any social media service to disable Singapore users' access to egregious content and stop the egregious content from being transmitted to Singapore users via other channels or accounts.
IMDA has, in fact, performed this role for some time now, working with social media services behind the scenes to deal with egregious content. As Singapore's media regulator, IMDA also has significant experience in assessing content across the different media platforms and making decisions to protect the community.
Under this Bill, IMDA will be better equipped to ensure Singapore users are protected from egregious content online. But IMDA will not have carte blanche to issue directions. Its powers will be limited in scope.
First, IMDA will not be able to issue directions in respect of private communications. Those will remain private.
Second, directions can only be issued for certain categories of egregious content relating to user safety.
The new section 45D proposed by the Bill defines "egregious content" to include content advocating terrorism, suicide and self-harm, violence including sexual violence, child sexual exploitation, content posing public health risk and content likely to undermine racial and religious harmony. These categories will be set out in law.
When dealing with content that requires the expertise of other agencies, IMDA will consult them accordingly. As an example, when assessing content pertaining to public health measures and risk, IMDA will consult the Ministry of Health (MOH) and its experts.
The new section 45M proposed in the Bill requires designated services to take all reasonably practicable steps to comply with an applicable Code of Practice.
If they do not, IMDA can take regulatory action under the proposed section 45N to issue (a) a financial penalty; or (b) a rectification direction requiring the service to remedy the failure to comply with the Code of Practice. Non-compliance with a rectification direction will be a criminal offence, punishable with a fine. For egregious content, non-compliance with a direction by IMDA will also be a criminal offence, punishable by a fine.
Mdm Deputy Speaker, I said right at the beginning that laws are necessary but success alone in ensuring our citizens' safety cannot just depend on the laws. Respondents of MCI's public consultation and engagements agreed with this view. They wanted the Government to mandate stronger measures and social media services to do more to reduce harmful online content. At the same time, they emphasised that all of us, as users of social media services, have an individual responsibility to protect ourselves.
During one of our engagements, Mr Mark Joel Premraj, a parent, shared his perspective on how parents also play a key role in educating their children on inappropriate content online, including how to encourage them to flag the inappropriate content they come across.
Besides establishing a robust regulatory toolkit, the Government has taken active steps to nurture a well-informed and discerning public. Efforts to educate the public include the National Library Board's S.U.R.E. programme. It equips the public to think critically, be responsible producers and consumers of information, and stay safe and well online. Since its launch in 2013, S.U.R.E. has conducted over 6 million physical and digital engagements.
In addition, the Ministry of Education (MOE)'s refreshed Character and Citizenship Education curriculum has a stronger focus on Cyber Wellness education, where students learn to be safe, respectful and responsible users of cyberspace, and to be a positive peer influence.
In support of the Digital for Life movement, launched in February last year, community partners have also spearheaded initiatives which helped over 270,000 Singaporeans enrich their lives through digital technologies.
For example, TOUCH Community Services has partnered Meta to conduct the Digitally Ready Families programme, where low-income families learn digital life skills and cyber wellness tips. Another Digital for Life partner is "Kids PlaySafer". Created and run by Ms Sandra Low, a mother of an 11-year-old and 9-year-old, "Kids PlaySafer" has conducted talks on digital literacy and cyber safety to help parents manage their children's digital needs.
Mdm Deputy Speaker, may I continue in Mandarin, please.
Yes, please.
(In Mandarin): [Please refer to Vernacular Speech.] Mdm Deputy Speaker, for many people, technology is indistinguishable from magic that greatly improves our lives and brings about greater convenience. However, for the parents of 14-year-old British girl, Molly Russell, social media became the dark spell that took away the life of their daughter. Badly affected by the thousands of contents about self-harm and suicide, young Molly ended her short life.
Besides content about self-harm, there are many other types of harmful online content including those that promote violence, sexual abuse or enmity between races. If we allow such content to flood our cyberspace, many people, especially our young, would be adversely affected; and the social cohesion that we built preciously over the years may also be at risk.
That said, there is no law that can shield us totally from harmful content. Therefore, as the Government strengthens the law, our hope is that as parents, we can encourage our children to tell us whenever they encounter problems online, so that we can support them. Our hope is for social media platforms to innovate and come up with newer technologies to protect users, beyond just fulfilling their obligations.
All of us also hope that while civil organisations and individuals strengthen their own awareness of such content, they can also empower vulnerable groups to counter such content.
Only then, can we move forward as one.
(In English): Mdm Deputy Speaker, let me conclude. When the author Arthur C Clarke published "2001: A Space Odyssey", one of the lines in the book became famous and quoted many times over. It says, "Any sufficiently advanced technology is indistinguishable from magic".
Mdm Deputy Speaker, magic happens as we speak. Without disturbing Parliamentary proceedings, Members can compare notes instantaneously, in fact, I saw some of you do so. And conduct research on the fly either directly by instructing the colleagues outside the Chamber or just looking at the QR code that I distributed.
Gone are the days where we might rush home to catch a favourite television programme. So much content is available online anytime, anywhere.
But not all of this content is good. To ensure that safety is upheld for Singapore users, we need OCSs to be held accountable. Equally, we need the support of everyone in the community to keep each other safe online. I appeal to Members to support this Bill so that we can together improve online safety. I beg to move. [Applause.]
Question proposed.
Mdm Deputy Speaker, I support the Online Safety (Miscellaneous Amendments) Bill.
When social media first emerged, many people saw it as an unadulterated good thing – a new frontier, separate from the real world, offering freedom of speech without ill consequences. Indeed, the founders of many social media companies had good intentions – to connect users from around the world; to give them personal control to express what they desire and exchange ideas; to be free of oppression; to come together and do good for themselves and their societies.
Years passed, and alas, we know that the reality is far from the ideal. We have seen how extremism gets propagated online, inspiring and triggering acts of terrorism in parts of the world. We have seen how young lives were lost to dark and depressing content, and irrational online movements such as the "Blackout challenge". We have seen how hate speeches, in the name of free speech, get disseminated and cause great divisions within society.
These are worrying trends that we see around the world and countries, including those that uphold democracy and free speech, are facing the same challenges. Singapore, too, is not spared. As a society, we must respond to curb its ills and protect the vulnerable, while allowing the widest possible.
So, our legislation must evolve to better protect our people from harmful online content, especially minors who are more susceptible and vulnerable. In this, Singapore is not alone in passing such a legislation. Australia and Germany are some of the first few countries to have done so, and others are either doing the same, or thinking of doing the same.
The Online Safety (Miscellaneous Amendment) Bill builds on some of our existing laws to further recognise that happenings online can have a real impact in the physical world and actions must be taken to address issues concerning the safety and well-being of our people. The Bill also represents a shift in some ways. Firstly, protecting users is no longer just the responsibility of the Government or the individual users themselves, but also the OCS providers, who will now be explicitly required to implement tangible measures or face legal consequences for failing to do so. Secondly, giving teeth to our agencies to compel these OCS providers to act on harmful and egregious content, even if these service providers are not situated on our shores.
On the Internet, all types of information, the good and the bad, are readily accessible and can be proliferated widely almost instantaneously, thereby extending the harm it can bring onto innocent people. In instances of bullying, the Internet magnifies the effects. And, in other instances, the widespread egregious content could mislead and cause harm to the innocent. Thus, I believe that most, if not all, will agree on the importance of protecting those who are vulnerable online.
Nonetheless, I have the following questions to ask about the Bill.
Firstly, I think it is quite clear that the Ministry is taking a whitelist approach to applying the Codes of Practice. Therefore, I would like to ask: how does the Ministry determine which OCS providers make it into the whitelist? What criteria or parameters do the Ministry consider?
Second, I note that private messages will not be covered in this Bill. I suppose the Ministry is trying to strike a balance between offering sufficient protection whilst not being overly intrusive. However, there could be instances where objectionable content gets shared through private messaging channels. Moreover, in the context of Australia where its online safety act was passed just last year and came into effect earlier this year, private messaging is included. Could the Ministry explain the considerations made when deciding what to cover and what not to cover? Perhaps, the Ministry could also share the lessons learnt or the observations made of other jurisdictions where similar legislation was passed before us.
Thirdly, while the Ministry has a broad list of what constitutes egregious content, such as content that advocates or instructs on suicide, self-harm, violence, child sexual exploitation, public health risk, racial and religious disharmony and terrorism, who and how will this assessment and decision be made on what actually crosses the threshold to qualify as egregious?
While I applaud the Online Safety (Miscellaneous Amendments) Bill for being a timely one, I would also like to know: how frequent does the Ministry intend to review the Bill? As we know, technological advancement and disruptions happen at accelerating speeds. Along with it comes new issues and operational challenges that the Bill today may not be adequate in addressing.
For example, with the rise of the metaverse and the increasing number of young people immersing in it, our legislation and protection mechanisms must catch up quickly. In May this year, it was reported on various news sites, such as the BBC and The Business Insider, that a researcher's avatar was sexually assaulted on a metaverse platform called the Horizon World. The researcher from a non-profit advocacy group SumOfUs entered the particular metaverse and, within an hour, her avatar was raped in the virtual space. As she wore her Virtual Reality (VR) equipment during that episode, her controller vibrated when the male avatar strangers touched her, resulting in a physical sensation consequential of what she was experiencing online. The incident left her feeling "disoriented". It was a clear instance of how the virtual and physical world boundaries have blurred and how online happenings can cause real harm.
Apparently, that was not an isolated incident, as there were other reports of similar sexual assaults, homophobic and racial slurs, as well as gun violence on Horizon World. Though these reports were specific to Meta's metaverse platform, it is not difficult to imagine similar incidents happening on other metaverse platforms, especially if the different metaverse universes start to connect with one another.
Therefore, how will the Bill, in its current form, be able to protect users from such online harms?
In addition, as we look to a possible future of Web 3.0, where data and control become decentralised, going just after a couple of OCS providers operating in the Web 2.0 world may not be effective enough. Hence, we will need to ensure that our legislation and enforcement capabilities are updated in a timely manner, so that while we do not want to be over-prescriptive, we are also not too big a step behind these technological developments.
Therefore, I hope the Ministry can also share more about what is being done to continually engage industry players and community stakeholders so that our legislation and Codes of Practice not only have teeth but will bite where it matters. With that, I support the Bill.
Mr Gerald Giam.
Mdm Deputy Speaker, the Online Safety Bill before us seeks to tackle harmful content on online services like Facebook, YouTube and TikTok, which are accessible to users in Singapore. I support the Bill, given the online harms that people in Singapore have been subject to on social media and on the Internet, and the growing need to protect our people, especially the young, from these harms.
However, I have some clarifications to seek on the Bill which I hope the Minister will address before we vote on the Bill.
Access to digital communication devices is not optional in this day and age, even for younger children. For example, if a 10-year-old child were to take public transport on his own to and from school, his parents would want him to be able to contact them in case of an emergency or to track his location. In most cases, this can only be done using a mobile phone or smart watch. However, it will be unwise to give that same 10-year-old unfiltered access to the Internet on his phone.
Currently, parents can install a parental control app on their child's phone. This app will allow parents to restrict content, approve apps, set screen times and filter harmful content. It can also locate the child using GPS.
I set this up for my son some time back. However, even with all my professional technical knowledge, it took me quite a bit of time and research to figure out which was the most suitable software to use and how to configure it. I wonder how many parents have tried to set up parental control software for their children. For those who have not, they should be aware that their children and teens essentially have unfiltered access to the Internet and all the harms that come with it. These parents can only regulate their children's Internet access by looking over their shoulders. This is a suboptimal solution, given the asymmetry of technical knowledge between most parents and their children. Most children nowadays can run rings around their parents when it comes to configuring settings on their mobile phones.
Also, for such content filtering to work for young people, age verification is needed. The Code of Practice for Online Safety for Designated Social Media Services proposed by the Ministry states that social media services must have additional measures to protect children, including minimising children's exposure to inappropriate content and ensuring that their account settings are age-appropriate by default.
However, the Code of Practice does not prescribe how this age verification should be implemented. Indeed, attempts at imposing age verification have previously failed in the United Kingdom's implementation of the Digital Economy Act of 2017, in part because of privacy concerns. Online age verification providers could collect excessive personally identifiable information and process it for other purposes in violation of privacy laws.
Separately, a young user can circumvent age restrictions by declaring his age to be 18 when, in fact, he is only 12. Can I ask the Minister: how will content providers be required to perform age verification checks in practice?
Some Internet service providers do provide parental control tools which block harmful content before they come through the fibre. However, they require a separate subscription that entails an additional cost each month. Many parents are not even aware of this service. They will have to take the effort to log in to their broadband provider's website and subscribe to this service. This additional friction will deter many parents from signing up, leaving young children vulnerable to accessing harmful content without their parents' knowledge.
It would be better for Internet service providers to block harmful content at the network level by default, rather than expect parents to set up complicated filtering software on their children's devices. This remote filtering should be activated by default for all new mobile and broadband subscriptions and offered for free for all subscribers. This will ensure that even children of less tech-savvy parents will be protected by default. Adults who need full access to the Internet should be able to opt out of the filtering service without any charge.
I am glad to note that under the Code of Practice, content that may encourage young users to engage in dangerous acts will be considered harmful content and be subject to additional safeguards for young users. Examples of these include the "Skull-breaker Challenge" where two people trick a friend standing in between them to take a vertical jump, then kick their legs from under them as they are in the air, making them fall backwards and potentially injure their head and back. People sometimes do not properly assess the risk associated with an activity. They may have seen others perform it without incident in a YouTube video and may be tempted to experiment themselves.
Ultimately, we cannot completely insulate young people from all dangerous, harmful and silly online content. The best protection is for parents and teachers to educate their children and students of the potentially harmful content that may be accessed online and the consequences of indulging in them. The Media Literacy Council could also directly push out educational materials on the platforms that young people access, like TikTok and Telegram. This should be an ongoing process, not a one-time effort, because harmful content is constantly evolving, and new trends are always emerging.
Under this Bill, failure to comply with the directions from IMDA could be an offence punishable by a fine on conviction. Can the Minister clarify if this fine will apply to only the company or also the individual officers within the company responsible for ensuring compliance? Given the financial might of social media companies, they might have no problem paying even a huge fine.
The Code of Practice will require social media services to submit annual reports to IMDA to reflect Singapore users' experiences on the service, including the actions that they have taken on user reports.
I would like to propose that social media services also be required to submit quarterly reports, listing the type of content that has been flagged by users. This is so that IMDA can be kept apprised of trends in harmful online content and behaviours.
Section B of the Code of Practice requires that users of OCSs must be able to report harmful content or unwanted interactions to the platform providers through an "effective, transparent and easy to use mechanism" and social media services are expected to take action on these user reports in a "timely manner".
This leaves open lots of room for interpretation. In contrast, Australia's Online Safety Act requires platforms to provide a clear and easily accessible complaints system for end-users to submit complaints or requests to remove certain material and the platforms must respond to the complainant within 48 hours, failing which the end-user may contact the eSafety Commissioner, who has the power to investigate the complaint. I would like to propose that Singapore's Code of Practice include these specific requirements and timelines.
Another potential area of harm to young people is online gaming, which can be both addictive and cause social problems.
I note that the Bill covers cyberbullying content that is likely to cause harassment, alarm or distress to the targeted person. Will the non-consensual sharing of intimate images be covered in this Bill? There have been cases, recently, of disgruntled ex-lovers sharing such images, which, most certainly, will cause alarm and distress to the victim.
Next, I would like to seek clarifications from the Minister regarding the protection of Singaporeans' democratic rights in this Bill. Some respondents to the public consultation sought assurance that the proposed measures would not affect user privacy or freedom of expression.
The Bill gives wide-ranging powers to IMDA to issue directions to social media companies to remove harmful content if it deems it so. Can the Minister elaborate on what safeguards will be in place to ensure that such powers are not abused? Will there be channels for independent appeal or judicial review?
The UK's Online Safety Bill specifically includes protections to safeguard pluralism and ensure Internet users can continue to engage in robust debate online. For example, section 29 of the latest draft of the UK's Online Safety Bill requires content providers to "have regard to the importance of protecting the rights of users and interested persons to freedom of expression within the law", when deciding on safety measures and policies.
Section 15 of the UK Bill also requires social media services to put in place clear policies to protect "content of democratic importance", such as user-submitted comments supporting or opposing particular political parties or policies, and to enforce this consistently across all content moderation. The UK Bill also requires that platforms must not discriminate against different political viewpoints.
The UK's draft legislation has also been designed to safeguard access to journalistic content. News publishers' content will be exempted from social media platforms' new online safety regulations. Because of this, social media platforms will not be incentivised to remove news publishers' content, as a result of a fear of sanction from the regulator.
Are there such provisions in Singapore's Online Safety Bill? If not, will the Government study the Online Safety Bills of other countries, including the UK and Australia, and include democratic protections in the Code of Practice and subsidiary legislation?
Will Singapore have the equivalent of an eSafety Commissioner like Australia does? Who will this eSafety Commissioner be, and will he or she be empowered to make directions independent of the Government?
Australia's Online Safety Act itself was controversial in part because of the huge amount of discretion and power it puts in the hands of the Minister for Communications and the eSafety Commissioner to determine what are community expectations. How will Singapore's Bill safeguard democratic freedoms while protecting the young from online harms?
I note that some electronic services are excluded from this Bill. Examples of these are SMS and MMS services. Can I confirm with the Minister that other private messaging platforms like WhatsApp, Telegram and Signal are also excluded from this Bill? For the avoidance of doubt, I am not advocating for these services to be included in this Bill, as they are primarily used for private communication between individuals. Much of the communication is end-to-end encrypted, which means even the platforms do not have access to the data exchanged by their users. I would have strong privacy concerns if this encryption were to be broken for the sake of enhancing online safety. Madam, I look forward to the Minister's responses.
Mr Zhulkarnain Abdul Rahim.
Mdm Deputy Speaker, I stand in support of the Bill. Digital technology has permeated our lives. It has a deep impact on how we learn, how we do business and how we interact with one another. However, because of the cloak of anonymity, the online world rears its ugly head through online harms.
Maintaining safety online is not just the responsibility of the individual or the Government but all stakeholders involved. In this regard, many felt that stricter enforcement of relevant laws can be effective to combat or reduce online harms. Many also felt that technology companies and platforms must lead the way in tackling this issue, alongside the Government and us fellow Singaporeans.
It is, thus, timely that we have this Bill to further help us safeguard Singaporeans against online harms.
Last week, my firm organised its annual thought leadership platform, the Dentons Rodyk Dialogue 2022, themed "Building a Safe and Inclusive Digital World Together: Vision and Transformation". During the keynote speech, Minister Josephine Teo explained Singapore's approach in regulation towards digital safety and inclusion by ensuring what she described as the "3As".
Firstly, accretive – building each step or measure one after the other in a calibrated approach. Second, agglomerate – pulling in partners and groups in our collective endeavour. Third, agile – being able to adapt through different emerging technology or disruptions.
To borrow from bonds credit rating parlance, that to me, is a triple-A rated approach – sound and sensible.
In dealing with online harms, it is important to take a calibrated and multi-stakeholder approach while keeping a close eye on emerging technologies, such as Web 3.0 and the metaverse. Although Bill is focused on platform and service providers, there is also a need to focus on the users, particularly, the victims of online harms.
In June 2021, I started an initiative called Defence Guild SG, a collaborative group of lawyers providing pro bono assistance to victims of online abuse or harms. We now have over 20 lawyers who volunteer pro bono to assist or advise victims who face sexual harassment or online harms. The bulk of the cases they face are time-sensitive given the viral nature of the harmful online content, but most importantly, they are emotionally draining and usually, it is time- and cost-consuming for victims to seek redress.
Following this initiative, together with other People's Action Party (PAP) Members of Parliament, like Ms Hany Soh and Ms Nadia Ahmad Samdin, we spearheaded the formation of resource toolkit to combat online harms at the PAP Women's Wing International Women's Day celebration in March this year. And in September this year, with the help of pro bono lawyers, social workers and counsellors, we ran a workshop for close to 70 activists across various PAP branches on the practical walkthrough of the resource toolkit. This is to support residents during Meet-the-People sessions or any of our other activities.
Notwithstanding all of the efforts, according to a recent survey, almost 57% do not know what are the legal redress or help that they can get when faced with online harms personally. Hence, raising awareness and empowerment is a continuing endeavour.
In this regard, I have three suggestions: first, to provide legal pro bono advice under the Legal Aid Bureau for individuals facing online harms, much like what Defence Guild SG is doing for victims currently; second, to standardise the ease of reporting of online harms across all platforms; third, to seek platform and broadcasting service providers to be included in public awareness campaigns on eradicating online harms. In Malay please, Mdm Deputy Speaker.
(In Malay): [Please refer to Vernacular Speech.] In June 2021, I started an initiative called Defence Guild SG.
This was in the wake of a harassment incident towards 17 local religious teachers, who became victims of a lewd online survey. This issue raised awareness among many within our community about the threat of online harms.
Over 20 lawyers from different races and religions, comprising of veteran and young lawyers from the Lawyers@M³ network, joined forces to provide legal advice to victims of online harassment.
Most of the cases they dealt with are sensitive in nature and many within our community are unaware of the type of help available as well as their own rights. However, this protection could not have come about without the partnership of those from the Government, individuals and technology companies or online service providers. Therefore, this Bill places responsibility in the hands of service providers, to block online content that may cause harm, such as extremist content, violence and those that negatively impact our multiracial and multi-religious society.
I, therefore, support this Bill.
(In English): I now move away from the individual and on to the service providers, the focus of this Bill.
I have several clarifications: firstly, I would like to clarify on the definitions used in the Bill; the second set of clarifications relate to the mechanism of the takedown and defence available; and lastly, I have some clarifications and suggestions in relation to the Code of Practice proposed.
First, on the definitions. In determining whether a broadcasting service is private or domestic, the new section 2(3) of the Broadcasting Act gave regard to certain factors, and I will touch on two of them.
Firstly, on the number of individuals in Singapore who are able to access the content. The clause itself does not state, for guidance, the actual number as a threshold. In this regard, may I ask what would the threshold be? Should it be a percentage of entire users of the service who are in Singapore? Some clarity in this regard would be helpful.
Secondly, on the restrictions on who may access the content. Would content on accounts which are accessible through friends or restricted followers, be considered "private or domestic nature"?
Next, in respect of the new clause 45D which sets out the definition of "egregious content", I have a few questions.
First, what is meant by content that "advocates or instructs"? Would the Ministry consider content to be harmful by looking objectively at its impact on our society, even though the content itself may not provide a clear set of instructions or advocate a certain position?
Second, on resources, may I ask who will be the enforcement agency for this? And would there be a dedicated team to regulate such behaviour? And if so, how are we sufficiently resourced for this?
For example, in Australia, the Department of Home Affairs has a dedicated team to find content on social media sites that promote hate, incite violence, or points to terrorist propaganda. The team has a budget of around AUD$3 million. It is a resource-intensive endeavour, and if we rely on self-reporting by individuals or self-regulation by providers, there may be many instances of online harms that fall through the crack.
Thirdly, I note that certain categories of harms are expressly stated in the Bill. May I ask what about issues relating to drug abuse or any other activities that are illegal or against our social norms – would these be considered harmful content as well? May I suggest that certain discretion be given to include any future definition of harmful content.
My next clarifications are on the mechanism. I welcome the takedown or disabling order under clause 45H and the blocking order under clause 45I. In particular, the new clause 45E makes it an offence of not stopping egregious content on an OCS.
There is also a defence available to service providers in not complying with the order, if it was not reasonably practicable to do more than what was in fact done, or if there was no better practicable means than what was in fact used.
However, may I ask what is meant by "reasonably practicable"? Would the costs or expenses involved in complying with such order be a relevant factor?
Likewise, what about actual or consequential loss that the party may suffer when complying with such duty? I would suggest that such costs, expenses and losses are not relevant factors. This is because economic losses or ramifications should not be placed on equal importance footing with online harms that have debilitating and irreversible effects on our society and individuals. Nevertheless, I would welcome Minister's clarifications in this regard.
Next, in relation to the immunity given to service providers against criminal or civil liability when complying with the orders – under clause 45J. Would this apply to civil liability brought by parties from outside the jurisdiction as well? I understand and fully appreciate that our laws may not have extraterritorial effect in this respect.
I am fully aware that the current Bill, as proposed, was done after extensive consultation with service providers and platforms and that they are supportive of such proposals. Other jurisdictions have also passed similar legislation to safeguard against online harms. Perhaps the Ministry can also explore reciprocal immunity provisions with other like-minded jurisdictions in the future. This would help harmonise and set an international standard for compliance of cross-border directions in the future.
My next clarification is in relation to clause 45H(1)(d) on the period of takedown notice. Can we consider a fixed period? For example, Germany's similar law provides for a 24-hour requirement for takedown. This reflects the imminent risk of such online harms. If there is a standard fixed period by legislation, it will lead to a reasonable expectation or standard within the industry for compliance and sufficient company internal compliance processes can be put in place to meet with such expectations. Alternatively, perhaps, Minister can clarify what is the estimated period that is envisaged for a typical direction or order?
Lastly, may I clarify what is meant by clause 45H(2)(b), that a requirement "must not require the doing of anything with respect to the provision of an online communication service to the whole or part of any area in Singapore"? Perhaps, can Minister provide an example for this?
Finally, Mdm Deputy Speaker, in relation to the Code of Practice, I welcome such Code of Practice. I understand that the a Code of Practice can be revised or amended pursuant to a process to be followed, which allows for future flexibility and adaptability.
My question is, given the fluidity and dynamic nature of our digital ecosystem, would the Code be able to keep up with the changes brought upon by the rapid changes in technology?
I note that there is a process before a change in Code can be made, as envisaged at clause 45L. However, how long would that take? Perhaps, if I may suggest the formation of a council to formulate and update the Code regularly to keep up with technology trends? Such Council can comprise of service providers, regulators, important law enforcement or community stakeholders.
In conclusion and most importantly, digital safety is the responsibility of everyone. Notwithstanding the above clarifications, Mdm Deputy Speaker, I stand in support of the Bill.
Mr Leon Perera.
Mdm Deputy Speaker, the Online Safety (Miscellaneous Amendments) Bill marks a step in the right direction to create some regime for enforcement of basic standards of protection and decency against acts of online harm. I support the thrust of the Bill and stand in agreement with the arguments made by my Parliamentary colleague, Aljunied Member of Parliament, Mr Gerald Giam.
In particular, I strongly support Mr Giam's call on the Government to provide clear and unequivocal assurances that this law will not be used to curtail the exercise of legitimate free speech that touches on the public acts of public figures and that is not of a vicious and personal nature. I hope the Government can provide such assurances during the Parliamentary debate, to be recorded in the Hansard. Members will recall how the Protection from Harassment Act (POHA) was used by the Ministry of Defence (MINDEF) to take action against an individual, the legitimacy of which was disputed in a subsequent ruling of the Court of Appeals.
Madam, my speech will focus on the topic of online bullying, particularly as it pertains to children and of the harms of loot boxes and other gambling-like elements in electronic games. Before I address these two topics, let me make some general suggestions and clarifications on the approach taken in this Bill, where more clarity is, perhaps, needed.
Madam, under the proposed Bill, there are two key parts to the regulatory approach. Firstly, requiring OCSs with significant reach or impact to comply with Code(s) of Practice; and secondly, dealing with egregious content on an OCS that is accessible by Singapore users, by enabling IMDA to issue directions to deal with such content.
In addition to such measures and perhaps embedded in the Codes of Practice, I wonder if there could be a system where users of OCSs can first report content on the platform that is egregious and/or seriously harassing, to the OCS itself. If the OCS fails to take action within a stipulated time frame, say seven days, then there could be a mechanism whereby the complainant can raise this issue to the IMDA and ask the IMDA to issue the appropriate directions in relation to such content.
This draws on the Australia's Cyberbullying Scheme. One of its features is that a person may make a complaint to eSafety about cyberbullying material that targets an Australian child and this acts as a safety net, because they must have first reported the material to the relevant online service provider before taking the step. In addition, if the OCS fails to investigate or take action about that content reported by a user within the stipulated time frame, there could be penalties for the OCS.
While there are already reporting platforms or tools on most, if not all, of the OCSs, such a requirement embedded in a Code of Practice would create some legal pressure on the OCS to further investigate and act on user notifications or reports of egregious content in a timely manner.
Next, section 45D defines "egregious content". While this section does contain some specific definitions of egregious content, there could be other types of egregious content where more specificity would be welcome, as my colleague Mr Giam and other Members alluded to.
In particular, I think that the definition of egregious content given here in the Bill does not adequately deal with the following two categories: firstly, revenge porn or unwanted sharing of intimate images; and secondly, cyberbullying.
But at paragraph 4 of the First Reading speech, the June 2022 survey by MCI found that, "Sexual content, cyberbullying and violent content were the top three types of content that respondents felt the young needed to be protected from most". And I was actually not referring to the speech, but to the statement from MCI.
It would seem that the proposed Bill might not address the first two concerns in a sufficiently specific manner. In Canada, there is an offence of sharing intimate images without consent. I hope the subsidiary legislation can be more specific about these two types of egregious content.
Next, Madam, I note that at paragraph 8 of the Ministry's statement. it is stated that "The Codes of Practice may require Regulated Online Communication Services (ROCS) providers to put in place measures on their services to mitigate the risks of danger to Singapore users from exposure to harmful content and provide accountability to their users on such measures."
I would like to ask the Government how this accountability will come about. Will the Government commit to public consultations when new Codes of Practice are issued? This is hinted at in section 45L(2), but it is not a requirement. I hope that, given the evolving nature of online harms, as well as the need to balance privacy and free speech concerns, that a proper consultation process will be the norm in future before new Codes of Practice are issued.
Next, Madam, part 9(d) of the same statement refers to how ROCS providers should collaborate or cooperate with the conduct of research studies by experts approved by IMDA. Such research would allow IMDA to understand the nature and level of the systemic risks in the ROCS, and the evolution and severity of such risks.
Madam, I support this provision. It is a very positive move for MCI to formally refer to the use of research in this manner, since this area of online harms is a rapidly evolving space and since more research needs to be done on the effects of certain online activities, particularly on children.
Mdm Deputy Speaker, I will now move to talking about online bullying of children. Madam, this is a serious issue in Singapore. A 2019 Programme for International Student Assessment (PISA) study found that 26% of Singaporean students reported being bullied at least a few times a month, compared to an Organisation for Economic Co-operation and Development (OECD) average of 23%. This is not a small number by any means.
Madam, this subject is somewhat personal to me, as there were short periods of time in both my children's primary school life when they were bullied by friends. It is traumatic for the child and can leave lasting psychological scars. But it is also hard on the parents, arousing feelings of concern, frustrations and yet, helplessness. Why do I say helplessness? Because sometimes, the parents feels that they cannot address the problem with a sledgehammer, by coming down hard on the bullies, who are themselves children and may not fully understand what they are doing.
The effects of cyberbullying can be deadly. In America, a girl called Megan Meier committed suicide three weeks before her 14th birthday and her suicide was attributed to cyberbullying on the social networking site MySpace. I think the Minister had referred to this example as well. Her classmate's mother had created a fake MySpace account, pretending to be a teenage boy, Josh. "Josh" messaged Megan on AOL Instant Messenger (AIM) saying something very hurtful. Megan killed herself shortly after.
Madam, page 9 of the Bill excludes certain services from the scope of the Bill, including SMSes and, in part (e), "an electronic service where the only user-generated content enabled by that service is communication between two or more end-users that is of a private or domestic nature." This would imply that the law excludes bullying that takes place among a group of children, where one or more children bullies one child on a chat group, on say the victim's Instagram page.
Madam, children are a group where cyberbullying can have a very serious effect, given their lack of maturity and lack of life experiences in accessing resources that could help them. A large-scale National Institute of Health study in the United States (US) found that "the child participants who experienced cyberbullying were more than four times as likely to report thoughts of suicide and attempts as those who did not." I think this subject of youth mental health, in the context of social media, has been the subject of much public discussion and much research and rightly so, as many experts are tending towards a view that there is a connection between the very extensive social media usage that we are seeing among young people, and the kinds of mental health issues that are coming to the fore.
Madam, how to deal with this very difficult question – and I confess that I do not have a legislative magic bullet to suggest here. I am not advocating for law enforcement agencies to police private conversations in a way that compromises privacy, could itself be subject to executive over-reach and could corrode our children's capacity to spontaneously interact with one another as well as learn social lessons thereby.
I would suggest that for now, the problem can be addressed through education and would like to call on MOE and the Ministry of Home Affairs (MHA) to explore this. There are helplines for young victims of bullying, including cyberbullying, such as Tinkle Friend by the Singapore Children's Society, which does good work and deserves support from all of us.
I am also aware that cyberbullying is addressed in the current curriculum relating to cyber wellness in schools. However, how effective has this been? I would urge the Government to conduct further research into this area and study innovative programmes that have worked around the world.
In particular, we need targeted education in primary and secondary schools that helps students to recognise that as bystanders, they have an important responsibility to step in and stop bullying, or at least to not cooperate and to flag out potential issues to teachers or others in authority positions.
Encouragingly, the PISA study I cited earlier found that 94% of Singaporean students agreed that it is a good thing to help students who cannot defend themselves. But does this translate to bystanders pushing back when bullying actually happens online? This needs to be studied. The role of bystanders here is crucial.
For this kind of education to work better, I am wondering if we can enlist students who have stood up to bullying, either as victims or bystanders, to be anti-bullying ambassadors who give talks in schools as to what they experienced, what they did, how they coped and also, when and how to bring the authorities in, as opposed to necessarily escalating every single situation to people in authority.
Order.
PAPERCUTS